Testing Heterolocalism: An Assessment of Latino Settlement Patterns in the Southeastern United States

Transcription

1 University of Tennessee, Knoxville Trace: Tennessee Research and Creative Exchange Masters Theses Graduate School Testing Heterolocalism: An Assessment of Latino Settlement Patterns in the Southeastern United States Kristian Dennis University of Tennessee - Knoxville Recommended Citation Dennis, Kristian, "Testing Heterolocalism: An Assessment of Latino Settlement Patterns in the Southeastern United States. " Master's Thesis, University of Tennessee, This Thesis is brought to you for free and open access by the Graduate School at Trace: Tennessee Research and Creative Exchange. It has been accepted for inclusion in Masters Theses by an authorized administrator of Trace: Tennessee Research and Creative Exchange. For more information, please contact trace@utk.edu.

2 To the Graduate Council: I am submitting herewith a thesis written by Kristian Dennis entitled "Testing Heterolocalism: An Assessment of Latino Settlement Patterns in the Southeastern United States." I have examined the final electronic copy of this thesis for form and content and recommend that it be accepted in partial fulfillment of the requirements for the degree of Master of Science, with a major in Geography. We have read this thesis and recommend its acceptance: Ronald Foresta, Shih-Lung Shaw (Original signatures are on file with official student records.) Anita I. Drever, Major Professor Accepted for the Council: Dixie L. Thompson Vice Provost and Dean of the Graduate School

3 To the Graduate Council: I am submitting herewith a thesis written by Kristian Dennis entitled Testing Heterolocalism: An Assessment of Latino Settlement Patterns in the Southeastern United States. I have examined the final electronic copy of this thesis for form and content and recommend that it be accepted in partial fulfillment of the requirements for the degree of Master of Science, with a major in Geography. Anita I. Drever Advisor and Major Professor We have read this thesis and recommend its acceptance: Ronald Foresta Committee Professor Shih-Lung Shaw Committee Professor Accepted for the Council: Carolyn R. Hodges Dean of the Graduate School (Original signatures are on file with official student records)

4 Testing Heterolocalism: An Assessment of Latino Settlement Patterns in the Southeastern United States A Thesis for the Masters of Science in Geography, The University of Tennessee, Knoxville Kristian Dennis August 2007

5 All maps, pictures, and graphs were originally produced by the author and may not be reproduced without express written permission. Copyright 2007 by Kristian Dennis The University of Tennessee All rights reserved. ii

6 ACKNOWLEDGEMENTS I would like to thank all of my family, friends, and the faculty of the University of Tennessee who supported me during a great time of personal growth and achievement. Your assistance and encouragement were vital to my success and helped me reach my goal of earning a Master s degree. Specifically, I would like to thank my advisor and major professor, Dr. Anita I. Drever, who has been invaluable in guiding me through graduate school by developing a well-balanced program of study in GIS, Economic, and Urban Geography and helping me develop the framework of this thesis in those disciplines of geography. Dr. Drever s enthusiasm, dedication, and practical approaches to population geography and how they positively impacted me, and this thesis, cannot be overstated. I also would like to thank Dr. Shih-Lung Shaw for continuing to expand my interest and knowledge in GIS and its application across various disciplines. I will be forever grateful for Dr. Shaw s willingness to share such great detail and his breadth of knowledge about GIS. To Dr. Ronald Foresta, thank you for sharing your wisdom about our urban world. Your ability and willingness to share numerous research methodologies in urban geography was of utmost importance in defining my graduate school experience and developing this thesis. To Dr. Bruce Ralston and Will Fontanez, thank you for assisting me in various research and work-related projects that will always influence my professional career. Also, thanks to everyone who gave me advice, shared their knowledge, or assisted me iii

7 during this thesis, especially for field research. Lastly, I must thank several members of my family whose support, encouragement, advice, wisdom, and love have helped me so much. Specifically: my grandmother Christine Kiki Jones, my parents Nolan Dennis and Nancy Dennis, my brothers Ron Earley and Richard Sterchi and sister Tracy Hammond, my aunts and uncles Cheryl and Jim Dickert and Cynthia and Al Thompson. Thank you, as well to everyone else in the family, for everything! iv

8 ABSTRACT Previous theories of immigrant integration indicate that spatial propinquity is a necessary ingredient for a cohesive ethnic community. Wilbur Zelinsky s heterolocalism theory suggests this is no longer the case in today s world where technology has drastically reduced the friction of distance in human interaction. This thesis uses a mixture of quantitative and qualitative techniques to test heterolocalism s applicability to emergent Latino communities in the Southeastern United States. The results of this research generally support Zelinsky s theory that a growing number of ethnic communities are socially but not spatially cohesive. v

9 TABLE OF CONTENTS CHAPTER 1 INTRODUCTION...1 THEORETICAL BACKGROUND AND PREVIOUS LITERATURE...3 ETHNIC ECONOMY...5 ETHNIC ENCLAVES...6 ETHNIC-CONTROLLED ECONOMY...8 SOCIAL FORCES...9 RECENT DEVELOPMENTS...10 CHAPTER 2 CITY SELECTION AND ANALYSIS...16 CITY SELECTION...16 TESTING HETEROLOCALISM...17 CHOROPLETH MAP ANALYSIS...19 DISSIMILARITY INDICES...28 CHAPTER 3 LATINO RESIDENTIAL AND BUSINESS PATTERNS...33 LOCATION QUOTIENT ANALYSIS...35 CHAPTER 4 SPATIAL ANALYSIS OF RESIDENCES AND BUSINESS...51 CHAPTER 5 ASSESSMENT OF LATINO COMMUNITY TIES...62 SURVEY METHODOLOGY...62 SURVEY DISCUSSION...65 LATINO MEDIA EVALUATION...67 LATINO BUSINESS CLIMATE...73 REMAINING HYPOTHESES...77 CHAPTER 6 CONCLUSION...78 POLICY IMPLICATIONS...80 REFERENCES VITA vi

10 List of Tables Table 2-1: Dissimilarity Indices for Latinos compared to whites and blacks in selected cities of the South...31 Table 3-1: Latino businesses located in block groups with a Latino residential location quotient greater than Table 4-1: Latino businesses by distance (A) inside, (B) 1 mile, (C) 2 miles, and (D) 5 miles from block groups with above average Latino population...57 Table 4-2: Percentage of businesses located in block groups with the highest Latino populations...61 Table 5-1: Survey results of how Latino community ties are maintained...63 Table 5-2: Latino media by city and category...68 Table 5-3: Latino businesses and services within city limits by category with population and dissimilarity indices (DI) vii

11 List of Figures Figure 2-1: Small, medium, and large cities in the South where Latinos are most concentrated...18 Figure 2-2: Location of selected cities in the South...19 Figure 2-3: Latino population by percentage in block group in small, medium, and large cities...20 Figure 3-1: Latino location quotients (LQ) by block group for medium and large cities...36 Figure 3-2: Latino population location quotients and business locations by block group to identify clustering...41 Figure 3-3: Latino businesses in redeveloped strip center on Nolensville Pike in Nashville, Tennessee...48 Figure 4-1: Medium and large city maps showing high Latino population by block group and Latino businesses by distance from those block groups...52 Figure 5-1: Revitalization of Collinsville, Alabama CBD due to Latino economic growth...76 viii

12 Chapter 1 Introduction Earlier theories of immigrant integration indicate that spatial propinquity is a necessary ingredient for a cohesive ethnic community. Wilbur Zelinsky and Barrett Lee s heterolocalism theory (1998, 2001) suggests this is no longer the case in today s world where technology has drastically reduced the friction of distance in human interaction. This thesis will be the first to test heterolocalism s applicability to emergent Latino communities in the Southeastern United States where to date there have been few studies of Latino immigration settlement patterns. It employs three primary techniques: statistical analysis, GIS analysis of residential and business location patterns, and field research. These are used in combination to discover if spatial propinquity at the neighborhood scale factors into the development of community ties within the Latino communities. Zelinsky and Lee noticed the ability of a growing number of ethnic communities to remain cohesive despite the fact that members were living apart from one another. In light of this, the authors offered a new theory for understanding the relationship between social and spatial integration called heterolocalism. Heterolocal literally means other or different place. Zelinsky defines heterolocalism as theory that intends to convey the possibility that an ethnic community can exist without any significant clustering, that is, when the members of a particular group are scattered throughout a city, metropolitan area or some larger spatial domain (Zelinsky 2001: ). There are six major characteristics of heterolocalism: 1

13 1. There is immediate or prompt spatial dispersion of heterolocal immigrants within the host country, or in this case, city. 2. Residence and workplace are usually widely separated, and, frequently, there is also a lack of spatial overlap between residence on one hand and shopping districts and sites of social activity on the other. 3. Despite the absence of spatial propinquity, strong ethnic community ties are maintained via telecommunications, visits, and other methods at the metropolitan, regional, national, and even international scale. 4. Heterolocalism is a time-dependent phenomenon. Although we can detect some partial manifestations in earlier periods, its full development is conceivable only under the socioeconomic and technological conditions established in the late twentieth century. 5. As is the case with other models such as assimilation and pluralism, heterolocalism can exist in both metropolitan and nonmetropolitan settings. 6. In contrast with the other models, heterolocalism has implications for the sociospatial behavior at the transnational, even global, scale. (Zelinsky 2001: 133) Based on these tenets of heterolocalism, it seems likely that the settlement patterns of Latinos in the South are heterolocal as there is little historical Latino settlement in the South and there are few existing ethnic enclaves to attract new arrivals. In fact, according to the Census Bureau, the Latino population in the Southeastern U.S has grown very rapidly, increasing 294% from 1990 to Further, many Latinos own vehicles and therefore do not need to live close to members of their community in order 2

14 to associate with them frequently. In order to test the theory that communities of social but not spatial propinquity are developing in the South, the thesis will explore the following questions: Are Latino residences dispersed spatially? Are Latino businesses also dispersed from one another and from Latino residences? If the answer to these questions is yes, then arguably the settlement patterns of Latinos in the Southeastern U.S. are best described as heterolocal. Theoretical Background and Previous Literature Historically, immigration studies in the United States have focused mainly on port-of-entry cities since most immigrant settlement occurred in those areas (Zhou, 1992; Suro, 1998; Waldinger, 1996, 2001; Logan, Alba, McNulty, 1994; Logan, Alba, et al, 1996, 1999, 2000). Cities such as New York, Chicago, Los Angeles, Miami, and San Francisco have received and still receive the greatest number of immigrants to the United States (Waldinger 2001). Up until the 1980 s, the majority of Latino immigrants to the United States moved to these cities. Beginning in the 1990 s, primary migration from other countries to the U.S. shifted away from the established port-of-entry cities. Secondary migration from port-of-entry cities to other areas further diffused Latino settlement patterns resulting in growing concentrations in the Southeastern United States (Davis, 2001; Frey and Liaw, 1999; Johnson et al, 1999; Munoz and Ortega, 1997). Since the early 20 th century, the relationship between social and spatial assimilation was primarily defined by the Chicago School s ecological approach (Park and Burgess, 1925; Wirth, 1927; Massey, 1985; Alba and Logan, 1993; and Alba et al, 1999). This approach states that immigrants initially settle in less desirable areas near the urban core of the new host city, thus forming urban ghettos or enclaves. Then, the theory 3

15 assumes that over time immigrants assimilate into the host city, both spatially and culturally, and this will be accompanied by the loosening of ties to their community and culture. Initially, new immigrant groups in the U.S. were expected to concentrate in slums crowded to overflowing with immigrant colonies, the second generation was expected to relocate to working-class districts with an ethnic character, and finally to aspire to assimilate into the Promised Land at the city s edge (Burgess 1925: 56). Massey (1985) used many Chicago School assumptions when he developed his spatial-assimilation model. Massey s spatial assimilation model explains the distribution of immigrant and ethnic populations across the urban landscape (space) of metropolitan areas. Spatial-assimilation has three key elements: space (urban landscape), social mobility (resulting from economic advancement), and acculturation. People s distribution across the urban landscape space is the starting point for determining interaction among people in Massey s model. In urban landscapes, people may live in high-density inner city neighborhoods, aging streetcar suburbs with medium density, or sprawling suburbs and these landscapes shape the interactions of their inhabitants, bringing them into contact with their neighbors. Social mobility due to economic advancement allows people to move to higher quality neighborhoods away from the city center, while developing broader social networks that operate in larger spatial scales than just the neighborhood level. Lastly, acculturation, which is the exchange of cultural values between ethnic groups, is reflected by the dispersed, suburban settlement patterns of an ethnic group. Of the three elements of spatialassimilation, social mobility from economic advancement is most important to understanding the development and dynamics of an ethnic economy. 4

16 Ethnic Economy Light and Gold define three types of agglomeration of ethnic businesses: the larger ethnic economy, the ethnic enclave economy, and the ethnic-controlled economy. The ethnic economy consists of ethnic entrepreneurs, ethnic employers, and ethnic employees regardless of their physical location within a city or region (Light and Gold 2000: 4). The ethnic enclave is a spatially concentrated community of ethnic businesses and residences (Wilson and Portes 1980). Portes and Bach (1985) extend the definition of the ethnic enclave economy beyond the ethnic self-employed or ethnic employees to include the locational cluster of firms whose owners and employees are coethnic and whose firms employ a significant number of coethnic workers (Light and Gold 2000: 14). For example, Cubans in Miami have an ethnic economy and an ethnic enclave economy due to the location of businesses and residences and the number of Cubans employed within Little Havana. Conversely, Mexicans have an ethnic economy in the Southwest U.S., but not an ethnic enclave economy due to dispersed business and residential locations and a lack of coethnic employment across the southwest (Light and Gold 2000: 14-15). The ethnic-controlled economy combines the ethnic economy, ethnic enclave economy, and ethnic niches while operating in the mainstream economy (Light and Gold 2000: 23). Ethnic niches are developed when ethnics are concentrated or specialized within an employment sector based on their ability to meet the demands of the labor market (Lieberson 1980; Waldinger 1996; Logan and Alba 1999; Wilson 2003). The development of ethnic economies is dependent on class resources and ethnic resources. Class resources are generally vocationally relevant material, such as property 5

17 and wealth, and cultural capital such as values, attitudes, knowledge, and skills. The specific forms of capital are financial, human (education and work experience), cultural (art, music, literature), and social (network) capital. Financial capital in the form of rotating credit associations is commonly used by Latinos across the United States in various ways for business and personal purposes. Ethnic resources are sociocultural and demographic features of the whole group that coethnic entrepreneurs actively utilize in business or from which their business passively benefits. Typical ethnic resources include but are not limited to, kinship and marriage systems, relationships of trust, ethnic-derived social capital, cultural assumptions, religion, native language fluency, a middleman heritage, entrepreneurial values and attitudes, rotating credit associations, an ideology of ethnic solidarity, and a generous pool of underemployed and disadvantaged coethnic workers (Light and Gold 2000: 102). Light and Gold contend that ethnic resources, in conjunction with class resources, contribute financial, human, cultural, and social capital to members of ethnic groups to help build an ethnic economy. The degree to which the spatial contiguity of these resources acts as a catalyst in an ethnic economy s growth remains disputed. Ethnic Enclaves Historically, ethnic enclaves have formed in port-of-entry cities as immigrants move to neighborhoods where members of the same ethnic group have settled. Spatial concentration of businesses and residences is critical to the formation of an ethnic enclave. The ethnic enclave commonly includes residences and meeting places but is foremost a community of businesses (Wilson and Portes 1980: ). An ethnic enclave is further defined as an ethnic economy based on business ownership by ethnic 6

18 group members (Logan et al 2000: 98). Typically, ethnic enclave communities contain industrial or commercial functions such as factories, retail, grocery, and restaurant establishments. Therefore, the addition of residences in an ethnic enclave offer agglomeration advantages economically and socially for an ethnic group. Ethnic enclaves offer inhabitants convenience in transportation and communication while also providing a level of security from the unfamiliar new host city. The ethnic enclave s economy (Wilson and Portes, 1980; Portes 1981) develops due to the spatial clustering of immigrant residences and many ethnic-owned businesses that employ co-ethnics. Vertical and horizontal integration of products along ethnic lines has been an important factor in the development of ethnic enclave economies as well (Light and Gold 2000: 12). Specifically, since members of various ethnic groups, through socialization, come to possess specific forms of knowledge and skill, ethnic communities have the ability to market their ethnic skills in order to obtain economic reward. For example, Jamaican immigrants often know about Reggae music and many Mexican immigrants can cook Mexican food. In American society, there is demand for Reggae music and Mexican cuisine. Therefore, each group can vertically integrate within an industry where they create, produce, distribute, market, and sell their respective products. Also, the development of ethnic enclave economies helps organize labor pools, since ethnic communities rely on networks to locate and fill jobs. Early migrants settled in ethnic enclaves both because they were forced to by law, and because they wanted the economic and cultural advantages associated with ethnic enclaves. Access to ethnic media is an example of a cultural advantage that became important in the development of ethnic neighborhoods. Book publishing, the recording 7

19 industry, and the film industry also indirectly contribute to the growth of enclaves. Ethnic food stores almost always distribute audio and video tapes, CDs, books, newspapers, and magazines (Light and Gold 2000: 183). Before the advent of the automobile, immigrants had to live close to stores to access resources. However, Kim (2000) points out this is no longer the case. He notes that many recent migrant populations lack the extent and geographical concentration typical of earlier arrivals. Kim credits advanced communication technologies with creating culturally cohesive but spatially dispersed ethnic communities: By informing geographically dispersed immigrants of community meetings and events, the media are the most powerful means of integrating and sustaining the community without the ethnic media, the nonterritorial community could not exist. (Light and Gold 2000: ). Ethnic-controlled Economy The ethnic-controlled economy is linked to the larger economy of the region and combines the ethnic economy, ethnic enclave economy, and ethnic employment niches. Ethnic niches are developed when ethnics are concentrated or specialized within an employment sector based on their ability to meet the demands of the labor market (Lieberson 1980; Waldinger 1996; Logan and Alba 1999; Wilson 2003). An ethnic economy is also defined as significant and persistent economic power exercised by coethnic employees in the mainstream economy (Light and Gold 2000: 23). This market power enables coethnic workers to influence hiring, wages, and job conditions to their own advantage (Light and Gold 2000: 46). Ethnic-controlled economies operate across a variety of scales through vertical or horizontal integration and concentrated or dispersed 8

20 economic activities from the neighborhood to national level. Ethnic groups that develop an ethnic-controlled economy in specific employment sectors, such as the Jewish in clothing manufacturing and Chinese or Mexicans in the retail and restaurant industries, are good examples of vertical and horizontal integration as well. Social Forces Massey and Denton (1993) argue ethnic enclaves differed from black ghettos in three fundamental ways. First, immigrant ethnic enclaves were never homogeneous and always contained a wide variety of nationalities. Secondly, most European ethnics did not live in immigrant ghettos. Lastly, ethnic enclaves became a transitory stage in the process of immigrant assimilation while black ghettos became a permanent feature of black residential life. For European immigrants, ethnic enclaves were places of absorption, adaptation, and adjustment to American society. They served as springboards for broader mobility in society, while blacks were trapped behind an increasingly impermeable color line (Massey and Denton 1993: 33). Consequently, ghettoization or spatial propinquity in some cases can negatively affect an ethnic group. Additional research by Massey and Denton of 1980 Census data for the thirty largest metropolitan with the highest Latino populations revealed Latinos (and Asians) had lower segregation rates compared to blacks. The dissimilarity index, which is a measure of evenness or distribution of a group in an area, for Latinos in 1980, was 49% compared blacks 75%. In fact, within most metropolitan areas, Latinos and Asians are more likely to share a neighborhood with whites than with another member of their own group (Massey and Denton 1993: 67). Despite their immigrant origins, Spanish language, and high poverty rates, Latinos are spatially more integrated in U.S. society than are 9

21 blacks (Massey and Denton 1993: 77). Furthermore, Latino and Asian segregation falls progressively as socioeconomic status rises while black segregation remains high regardless of socioeconomic status. In other words, according to Massey and Denton more recent immigrant groups such as Latinos and Asians are less likely to suffer from ghettoization, more likely to improve their status in society, and have greater freedom in selecting where they live when compared to blacks. Further research by Drever (2004) of ethnic groups in Germany confirms that space is much less influential on ethnic communities than previously thought. Also, Drever s research concludes that minorities living in ethnic neighborhoods were no more likely to feel isolated from goods and services, no more likely to feel isolated from German society, or to have a closer connection with the culture of their country of origin (Drever 2004: 1436). The importance of ethnic capital in social mobility and qualitative effects of neighborhood composition is demonstrated by Borjas (1999). Borjas states that an individual s ethnic background influences the process of social mobility. Ethnic capital, which is the human capital an ethnic group possesses, provides paths of employment and social contacts in the community that contribute to the upward or downward mobility of an individual. Therefore, ethnic capital acts as a positive or negative magnet or glue within the ethnic community. The spillover effects of ethnic capital within an ethnic group transmit both good and bad socioeconomic characteristics. These spillovers are amplified by ethnic concentration, yet Borjas points out that an individual s skill level and socioeconomic class are more important to social mobility than ethnicity. 10

22 Recent Developments According to the Census Bureau, the Latino population of the South, excluding Florida and Texas, increased 294% from 593,181 to 1,744, 865 persons between 1990 and The primary factors contributing to this growth are the availability of jobs in the South s expanding economy and social and employment networks that have connected people to these jobs. Latino immigrants to the South initially worked in areas of agriculture and food cultivation, such as poultry or pork processing, mushroom or tomato cultivation, and tobacco harvesting. More recently, Latinos have moved into manufacturing, construction, service, and professional employment such as restaurant, hotel, wholesale and retail trades, finance, insurance, real estate, legal, and medical services (Kochlar et al 2005: 23-24). In terms of nativity and ethnicity, 57 percent Latino immigrants in Alabama, Arkansas, Georgia, North Carolina, South Carolina, and Tennessee are foreign-born and 73 percent of those foreign-born Latinos are of Mexican descent, compared respectively to 41 percent and 64 percent nationally (Kochlar et al 2005: 8-14). Furthermore, Georgia, Tennessee, Arkansas, and North Carolina all have significant Latino populations that are U.S. born, which indicates internal migration from primary Latino population centers such as New York, Chicago, Miami, or the Southwestern United States. Research by Alba and Logan et al (1996, 1999) addresses emerging trends in suburbanization, spatial assimilation, and housing characteristics of ethnic groups. Their research of family and household status, socioeconomic characteristics (income and education), and immigrant-related characteristics (ability to speak English, nativity, time in U.S., etc.) confirms that recent immigrants seem much more inclined to settle in 11

23 458).. Furthermore, research on internal migration of U.S born and foreign-born Latinos and Asians by Frey and Liaw (1999) determined that, even though high-immigration metros still receive most new immigrants, education level affected whether or not Latinos or Asians moved outside of those metros to other cities, metropolitan areas, or states. Additional research by Johnson et al (1999) provides insights into newly emerging Latino communities through spatial analysis of settlement patterns and inmigration flows. Their analysis identified states, such as Tennessee and Georgia, which experienced rapid Latino population growth during the early 1990 s and Latino population concentrations within those high-growth states at the county and Metropolitan Statistical Area (MSA) level. Identification of Latino migration flows for selected MSAs revealed Latinos were moving to those new destinations from many different places. Heterolocalism has been used in research of immigrant and refugee groups in Portland, Oregon (Hardwick and Meacham 2005). This research verifies elements of heterolocalism by confirming that during their early years of settlement, immigrants are held together by a set of region-wide linkages that are independent of settlement patterns. At this larger scale of analysis, refugees and other migrants, united by cohesive ethnic, cultural, and social networks, travel from one community to another on interstate highways to shop at ethnic groceries, eat in ethnic restaurants, and attend regional church conferences and retreats, social events, and family gatherings (Hardwick and Meacham 2005: 555). The results of these studies reveal that there has been an evolutionary shift from ethnic enclave formation to spatial dispersion of immigrants both at the national and local scale in the United States. Historical settlement patterns along with recent influxes of 12

24 immigrants have resulted in six major types of U.S. immigrant gateways : former gateways (Cleveland), continuous gateways (New York, San Francisco), post-wwii gateways (Miami, Los Angeles), emerging gateways (Atlanta, Washington, D.C.), reemerging gateways (Denver, Seattle), and pre-emerging (Charlotte, Raleigh-Durham) (Singer 2004: 5). Pre-emerging gateways are defined as areas with few immigrants in 1980 that experienced sudden, very rapid growth in their foreign-born population during the 1990 s. Newly emerging gateways such as Atlanta and Washington, D.C. are areas that experienced rapid growth of foreign-born and native-born population between 1980 and In addition, it should be noted that by 2000, more immigrants in metropolitan areas lived in suburbs than cities, and their numbers were increasing more rapidly in the suburbs as well. Mike Davis also utilizes the 2000 U.S. Census Bureau to examine the growth and settlement of the Latino population in the United States. Davis points out that surprisingly little attention has been focused on the historical geography of Latino settlement patterns in nonborder cities (Davis 2000: 49). Davis contends that as emergent Latino pluralities and majorities outgrow the classic barrio, they are remaking urban space in novel ways that cannot be assimilated to the earlier experiences of either African-Americans or European immigrants (Davis 2000: 49). For example, transnational suburbs of cities have developed revealing new patterns of chain migration and settlement. Davis states that economic and cultural umbilical cords now permanently connect hundreds of Latin American and Caribbean localities with counterpart urban neighborhoods in the United States (Davis 2000: 96). Telecommunications, cheap airfares, and the public presence of Latino radio and TV programming in the U.S have 13

25 contributed to this recent development. Hernandez-Leon and Zuniga s (2002) research of Dalton, Georgia demonstrates how social capital has catalyzed Latino s movement from primary concentrations of settlement to new regions. Furthermore, their study shows that substantial numbers of new arrivals to southern states are secondary internal migrants, coming from large historical concentrations of Mexican immigrants, such as Los Angeles, Chicago, and Houston. Like Davis, Hernandez-Leon and Zuniga assert that the effects of Latino migration in Georgia, North Carolina, Tennessee, and Alabama have shifted the demographic and social boundary in the opposite direction (north), turning many towns and small cities in those states into border communities (Hernandez-Leon and Zuniga 2002: 3). Munoz and Ortega (1997) reached a similar conclusion in their comparisons of Latino settlement in the major regions of the U.S. The South was particularly noted for its lack of any Latino ethnic group concentration and the lowest percentage of Latinos living in urban areas (60.8%) of any region of the United States, which is due to their employment in agriculture. Analysis of Latino migrants to Charlotte by Smith and Furuseth (2004) shows Latinos have moved directly to maturing suburbs, outside the urban core. They found the availability of affordable housing was the primary determinant of their residential location decisions. As evidenced by recent research, the historical theoretical approaches to immigrant settlement do not adequately describe the settlement patterns of the Latino population in the South. Settlement in the urban core and ethnic enclave formation appear to be a relic of the past. More recent research reveals not only that the Latino population 14

26 in the South has grown rapidly, but that Latinos are settling in suburban areas of cities, metropolitan and nonmetropolitan areas, and smaller cities throughout the United States. My research is the first to test the heterolocalism theory on a large region and gauge its applicability in the American South. Research by Drever (2004) and Hardwick and Meacham (2005) proves ethnic community ties are being maintained and further developed while dispersed population settlement and spatial integration is occurring on a metropolitan level. Likewise, Smith and Fureseth (2004) and Davis (2000) prove Latinos are increasingly settling in suburban areas and enclave formation appears unlikely. My thesis seeks to more effectively assess the emerging Latino population settlement and business or service development patterns in the Southeastern United States on a regional and city-by-city basis to determine if social and cultural ties are being maintained in the spatially dispersed Latino population in the American South. 15

27 Chapter 2 City Selection and Analysis City Selection For this study, I have chosen to analyze the cities and towns proportionally most heavily impacted by Latino migration in the Southeastern U.S. based on data from the 2000 U.S. Census. I limited my study area to the states of Alabama, Arkansas, Georgia, Kentucky, Louisiana, Mississippi, North Carolina, South Carolina, Tennessee, Virginia, and West Virginia. Florida, Texas, and the metropolitan Washington D.C. area are not included because they have historically been ports-of-entry for Latino immigrants. Because one of the tenets of heterolocalism is that it operates in urban areas of varying scales (Zelinsky and Lee, 1998), I have also broken down my study sites by size based on the Census Bureau s definitions for Places, Urban Clusters, and Urbanized Areas. A Census Place is a concentration of population either legally bounded as an incorporated place, or identified as a Census Designated Place (CDP), which is an unincorporated community. For this study, incorporated places (cities) were included, while unincorporated CDP s were not included. An Urban Cluster is defined as a densely settled territory that has at least 2,500 people but fewer than 50,000. Lastly, an Urbanized Area is an area consisting of a central place and adjacent territory with a general population density of at least 1,000 people per square mile of land area that together have a minimum residential population of at least 50,000 people. Based on these definitions and the range of population for all cities in the study area, I have developed three population categories: census places with a population of 16

28 less than 50,000 (Small cities), populations of 50, ,000 (Medium cities), and populations over 250,000 (Large cities). In each of these categories, I have chosen to analyze the five incorporated cities with the highest percentage of Latino population, i.e. the cities most heavily impacted by Latino migration and where Latinos are most concentrated (see Figure 2-1). The small cities I will examine are Siler City, North Carolina, Collinsville, Alabama, Biscoe, North Carolina, Bells, Tennessee, and Monroe, North Carolina. The medium cities are Marietta, Georgia, Kenner, Louisiana, Rowell, Georgia, Jacksonville, North Carolina, and Sandy Springs, Georgia. The large cities include Charlotte, North Carolina, Raleigh, North Carolina, Nashville, Tennessee, Atlanta, Georgia, and Virginia Beach, Virginia. The location of these cities is generally dispersed throughout the South (See Figure 2-2). Jacksonville, North Carolina, Kenner, Louisiana, and Virginia Beach, Virginia are coastal cities, while all of the other cities are located inland along major transportation routes. Furthermore, two-thirds of the cities are located in metropolitan areas. Siler City, Collinsville, Biscoe, Bells, and Jacksonville are located in nonmetropolitan areas. All of the cities have strong agricultural, manufacturing, or construction industries. Major U.S. military bases are also present in three cities: Marietta, Jacksonville, and Virginia Beach. Testing Heterolocalism The first three hypotheses of heterolocalism provide the framework for my analysis. The first hypothesis of heterolocalism suggests there is prompt spatial dispersion of immigrants in a city or region. One way to test if this is happening in the South is to map the settlement patterns of Latinos. For each of the cities selected 17

29 Small cities (population < 50,000) where Latinos are most concentrated Percent Latino 39.33% 40% 30% 23.48% 23.24% 22.80% 20% 21.39% 19.93% 19.41% 19.37% 19.03% 19.02% 10% 0% Siler, NC Collinsville, AL Biscoe, NC Bells, TN Monroe, NC Asheboro, NC City Hardeeville, SC Saluda, SC Sanford, NC Selma, NC Medium cities (population 50, ,000) where Latinos are most concentrated. Percent Latino 20% 16.93% 13.62% 15% 10.61% 10.05% 9.93% 8.64% 8.56% 10% 7.80% 7.43% 7.25% 5% 0% Marietta, GA Kenner, LA Roswell, GA Jacksonville, NC Sandy Springs, GA Winston-Salem, NC City Durham, NC Concord, NC Burke, VA Metairie, LA Large cities (population < 250,000) where Latinos are most concentrated. 8% 7.36% 6.99% Percent Latino 6% 4% 2% 0% 4.72% 4.49% 4.18% 3.29% 3.06% 2.97% 1.86% Charlotte, NC Raleigh, NC Nashville, TN Atlanta, GA Virginia Beach, VA Lexington, KY New Orleans, LA Memphis, TN Louisville, KY City Figure 2-1: Small, Medium, and Large cities in the South where Latinos are most concentrated. Source: 2000 U.S. Census. 18

30 Figure 2-2: Location of selected cities in the South. Source: U.S. Census Bureau. above, I have created choropleth maps of the Latino population in each city utilizing ArcGIS 9.1 and Census 2000 data (See Figure 2-3). These maps show the Latino population by block group as a percentage of the total Latino population for each city. Second, I calculated the index of dissimilarity for each city to measure the degree to which the Latino population is spatially segregated. Choropleth Map Analysis Examination of the small city choropleth maps reveals that Latinos can be found in all block groups of each city. Latinos appear somewhat concentrated in the cities of Bells, Biscoe, and Collinsville, where slightly more than half of the Latino population lives within one block group. The next most populous block groups in these cities contain approximately one-third of the Latino population. These three cities have the smallest 19

31 Figure 2-3: Latino population by percentage in block group in small, medium, and large cities. Source: 2000 U.S. Census. 20

32 Figure 2-3: Continued. Source: 2000 U.S. Census. 21

33 Figure 2-3: Continued. Source: 2000 U.S. Census. 22

34 Figure 2-3: Continued. Source: 2000 U.S. Census. 23

35 Figure 2-3: Continued. Source: 2000 U.S. Census. 24

36 Figure 2-3: Continued. Source: 2000 U.S. Census. 25

37 populations and have only three or four block groups apiece. For Monroe and Siler City, the Latino population is noticeably more dispersed throughout multiple block groups across each city. In the medium sized cities, the Latino population patterns tend to be dispersed as well. In Jacksonville, NC the majority of Latinos live in four south sector block groups that are near or part of the Camp LeJeune military base. However, because the town s geography is heavily influenced by the location of the military base and by the desirability of waterfront property, the Latino population is moderately dispersed because the town itself is polycentric. The next most populated area of Jacksonville for Latinos is the far north sector, while the central sector is the least. Likewise, Kenner s Latino population is dispersed but largely concentrated on one side of the city. Kenner does have one block group that contains 14% of all Latinos in the city, while there are six other block groups that each contain between five to nine percent of Kenner s Latino population. It is important to note the impact of Hurricane Katrina and the ensuing migration will likely have changed this geography. In Marietta, Latinos are strikingly dispersed throughout all areas of the city. Like Jacksonville, Marietta has a military base and the block groups with the most Latinos generally form a northern arc around Dobbins Air Force Base from the southwest to the southeast part of the city. Latino settlement in Sandy Springs is also dispersed. Latinos are clustered both in the southern and northern part of the city, principally around the major commercial corridor in the city. However, Sandy Springs northern and Marietta s eastern neighbor, Roswell, displays the most concentrated pattern of Latino population settlement. The 26

38 block groups with the highest percentages of Latinos in Roswell are all located in the central sector of the city and are contiguous to one another. Latino populations are also fairly dispersed in the large cities. Atlanta, Charlotte, Nashville, Raleigh, and Virginia Beach have several block groups where Latinos are concentrated. Furthermore, those block groups are all located in the inner and outer suburban rings of these cities, not in the central core or inner city as has traditionally been the case for ethnic enclaves in gateway cities. In Atlanta, there is a block group in the northeastern part of the city where Latinos are highly concentrated. This area is adjacent to a suburban multi-ethnic enclave international corridor" (Walcott 2002) outside the city limits (North Atlanta CDP). However, there are four other distinct, suburban nodes of Latino settlement scattered across the city. Charlotte also has residential clusters of Latinos scattered across the city. Latinos can be found in the northern, southern, and eastern parts of the city; however they are most concentrated in the east, along Central Avenue. In other parts of the city, Latino residences can be found near commercial corridors. Generally, Latinos are concentrated in particular block groups, but these block groups are dispersed throughout Charlotte. This pattern is also supported by the research of Smith and Furuseth (2001). In contrast to Atlanta and Charlotte, the majority of Nashville s Latino population has settled in one part of the city. The southeastern part of the city contains several medium and high Latino population block groups. These block groups are spread along three commercial corridors that extend from the urban center to the suburban periphery. Therefore, although Latinos are concentrated in the southeast part of Nashville, they are spread out within this area. 27

39 The most dispersed Latino population patterns can be observed in Raleigh and Virginia Beach. Raleigh has moderate to highly concentrated Latino block groups scattered throughout except in the western and central city neighborhoods. Virginia Beach is an entirely suburban city with no clearly defined core. Hence, the Latino population, like the rest of the population, is scattered throughout the entire city. Based on this analysis, I conclude that the overall pattern of the Latino residences in these cities is dispersed. Further, in all cities- except the three smallest cities (Bells, Biscoe, and Collinsville) - Latinos generally reside in suburban areas. The three smallest cites as well as Atlanta, Nashville, and Roswell are the only cities where there are distinct concentrations of Latinos. However, Latinos in Atlanta and Nashville and in the three smallest cities are living outside these concentrations as well. In all of the remaining cities (Charlotte, Jacksonville, Kenner, Marietta, Monroe, Raleigh, Sandy Springs, Siler City, and Virginia Beach) Latinos are moderately to highly dispersed throughout each city. Dissimilarity Indices Another method of testing the spatial dispersions of Latinos is to calculate the index of dissimilarity (DI) for the Latino and White as well as the Latino and Black populations. The index of dissimilarity is a measure of evenness, or distribution of a group in an area. The dissimilarity index is considered the standard measure of segregation (Duncan and Duncan 1955; Taeuber and Taeuber 1965; Massey and Denton 1988, 1993). It measures the percentage of a minority population in an areal unit that would have to move in order for that population to be evenly distributed throughout an entire geographic area. The index values range between 0 (complete integration) and

40 (complete segregation) and are interpreted as a percentage. For example, a value of 0.72 indicates that 72 percent of the subject population would have to move to be evenly distributed in a city. In this case, the areal unit used is a block group with the geographic area being an entire city. I have calculated the index of dissimilarity according to the following formulas for the Latino-White and Latino-Black populations of each city: D. 5 1/ L w / W Where l = the Latino population of the block group L = the total Latino population of the city w = the White population of the block group W = the total White population of the city And D. 5 1/ L b / B Where l = the Latino population of the block group L = the total Latino population of the city b = the Black population of the block group B = the total Black population of the city. Since the majority of Latinos living in the South arrived after 1990, the Census 2000 data captures their initial settlement patterns. The dissimilarity indices I have calculated indicate that prompt spatial dispersion of Latinos has indeed occurred. Latinos in most cities tend towards integration rather than segregation. Analysis of the Latino-White DI 29

41 shows that 12 of 15 cities have DI values less than 50%, with eight of those cities in the 20-40% range (See Table 2-1). Latinos appear relatively integrated in comparison to the 49% dissimilarity index calculated for the 30 metropolitan areas with the highest Latino population from the 1980 Census (Massey and Denton 1993: 67). Furthermore, these results show that city size does not necessarily influence where Latinos are settling. In each of the city size categories, four out of five cities have dissimilarity index scores less than 50 percent. More impressive, Collinsville, Jacksonville, Kenner, Charlotte, and Virginia Beach all have Latino-White dissimilarity scores of 30 percent or lower. These scores indicate a general tendency towards integration and dispersed Latino population settlement patterns. Atlanta, Monroe, and Marietta all have scores between 52 and 61 percent and even these figures are not very high. The patterns for the Latino-Black dissimilarity indices are similar to those observed among the Latino-White indices. 12 of 15 cities have Latino-Black dissimilarity scores below 50 percent and half of the cities have dissimilarity scores 30 percent or lower. These Latino-Black dissimilarity scores indicate Latinos are living in the many of the same neighborhoods as blacks. Nine cities have lower Latino-Black dissimilarity scores than Latino-White dissimilarity scores, while seven cities have higher scores. A city with lower Latino-Black than Latino-White dissimilarity scores indicates Latinos are more integrated in black neighborhoods than white neighborhoods. While a city with lower Latino-White than Latino-Black dissimilarity scores indicates Latinos are more integrated in white neighborhoods than black neighborhoods. Of the seven cities with a higher Latino-Black dissimilarity index, Atlanta has the highest Latino segregation from blacks at 67 percent, while Nashville is second at 58 percent. It should be noted is 30

42 Table 2-1: Dissimilarity Indices for Latinos compared to whites and blacks in selected cities of the South. Source: 2000 U.S. Census. Index of Dissimilarity for Latinos Small City Latino-White DI Latino-Black DI Bells, TN Biscoe, NC Collinsville, AL Siler City, NC Monroe, NC Medium City Jacksonville, NC Kenner, LA Marietta, GA Roswell, GA Sandy Springs, GA Large City Atlanta Charlotte Raleigh Nashville Virginia Beach, VA

43 that Atlanta has a majority black population, while Nashville, Charlotte, Raleigh, Virginia Beach, Marietta, Jacksonville, Marietta, and Monroe are close to 30 percent black. Again, the Latino-Black and Latino-White dissimilarity indices reveal that Latinos are typically dispersed in Southern cities. Atlanta and Nashville are the only cities displaying Latino residential concentration and a tendency towards residential enclave formation. In sum, the analysis of choropleth maps and dissimilarity indices support the hypothesis that there is in fact prompt spatial dispersion of the newly arrived Latino population within new destination cities in the U.S. South. 32

44 Chapter 3 Latino Residential and Business Patterns Heterolocalism s second hypothesis is that residence and workplace are usually widely separated, and, frequently, there is also a lack of spatial overlap between residence on one hand and shopping districts and sites of social activity on the other (Zelinsky 2001: 133). The temporary and shifting nature of construction and agriculture employment ensures many Latinos do not live close to their workplace (Kochlar et al 2005: 23-24). Previous research by Ellis et al (2004: 626) indicates immigrants in other professions also live in neighborhoods outside of where they are employed. Furthermore, a service-driven economy (and municipal zoning) dictates workplaces are generally separate from residences (Zelinsky 2001: 138). For this analysis, I focus on ethnic shopping and social spaces, as opposed to employment locations, to determine the degree of overlap between Latino residences and businesses. I built a complete database of all Latino-oriented businesses and services with complete addresses that I found within the city limits of the Medium cities (Jacksonville, Kenner, Marietta, Roswell, and Sandy Springs) and Large cities (Atlanta, Charlotte, Nashville, Raleigh, and Virginia Beach) using phonebooks, chambers of commerce, newspapers, and field research. I used this database to identify, through geocoding/address matching in GIS, where Latino businesses and services are located and their proximity to the Latino population. It is important to keep in mind that since addresses are located on both sides of a street, a business on one side could be in one 33

45 block group while a business directly across the street is in another. Often, commercial zones or thoroughfares split residential areas and do not necessarily overlap only one block group polygon. For instance, if a commercial corridor was completely within the boundaries of a block group, the businesses along the corridor would overlap or be completely contained by that block group, not located only on the edge of a block group. Most of the businesses in this research are clearly located along block group boundaries, indicating the block groups are often, but not always, split along major road centerlines. I did not examine small towns because almost by definition, Latinos in smaller communities live close to one another and the spaces where they work, shop, and socialize. However, it is important to note that several small cities (Bells, Biscoe, Collinsville, and Siler City) have benefited greatly from many Latino businesses being established in once vacant central business district (CBD) storefronts. Further, I utilized location quotients (LQ) to identify residential clusters of Latinos at the block group level. I calculated location quotients for the Latino population in each of the census block groups in the medium and large cities to identify the areas more proportionally Latino than the city average. The location quotient is calculated as follows: LQ ( Bi B) ( Ci C) Where Bi = Latino population (i) in Block Group B = Total Block Group Population Ci = Latino population (i) in city 34

46 C = Total population in city. A location quotient of 1.0 means the census block group is the same percentage Latino as the city at large. A location quotient greater than 1.0 indicates that Latinos are overrepresented in the census block group. A location quotient value of 2.0 indicates twice as many Latinos are in a block group than is average for the city as a whole, for example. I produced the following graphs of location quotients to display the patterns of concentration of the Latino population for each city (See Figure 3-1). These graphs show the total range of location quotient scores for all block groups. The location quotients are all graphed using the same scale so as to enable inter-city comparisons. The location quotient benchmark value of 1.0 is redlined on the graphs. Location Quotient Analysis The majority of cities show the same location quotient pattern. The graphs reveal that most cities have several block groups with high location quotients (LQ > 2.0). Jacksonville, NC is the only city without a high (LQ > 2.0) Latino location quotient block group and Kenner only has one. This pattern is important because it shows that Latinos are concentrating in multiple locations, which produces an overall dispersed pattern of population settlement. The highest degree of dispersion is most evident in the Virginia Beach location quotient graph, as only a few block groups have location quotients over 2.0. These scores further validate that Latinos are settling in cities in a dispersed pattern. Additionally for this analysis, I looked specifically at census block groups with Latino residential location quotients greater than 1.0 in order to determine whether or not Latino businesses tend to cluster in these block groups. I produced a series of maps displaying block groups with location quotients greater than 1.0 and business locations. 35

47 Jacksonville LQ LQ Block Groups Kenner LQ LQ Block Groups Marietta LQ LQ Block Groups Figure 3-1: Latino location quotients (LQ) by block group for medium and large cities. Source: 2000 U.S. Census. 36

48 Roswell LQ LQ Block Groups Sandy Springs LQ LQ Block Groups Atlanta LQ LQ Block Groups Figure 3-1: Continued. Source: 2000 U.S. Census. 37

49 Charlotte LQ LQ Block Groups Nashville LQ LQ Block Groups Raleigh LQ LQ Block Groups Figure 3-1: Continued. Source: 2000 U.S. Census. 38

50 Virginia Beach LQ LQ Block Groups Figure 3-1: Continued. Source: 2000 U.S. Census. 39

51 By displaying the location quotients in this manner, the block groups where Latinos are over-represented are highlighted. Geocoding (address matching) of the Latino business or service locations allowed me to perform an intersect in ArcGIS to determine how many businesses are located in high Latino location quotient block groups, which I will discuss later in this section. The population and business locations in this series of maps (See Figure 3-2), while displaying overlap with block groups where Latinos are overrepresented, are generally dispersed throughout the cities. Two patterns emerge from this analysis: First, the Latino business and service locations are generally located along major highways that are suburban commercial corridors mainly composed of small to medium-sized strip shopping centers and malls. This pattern is identifiable in the maps by the linear strings or strips of Latino businesses in the cities. Second, while most businesses appear to be located inside or in close proximity to high location quotient block groups, there are many Latino businesses not located in the high location quotient block groups. Both patterns are attributed to the generally dispersed settlement of Latinos in suburban areas. Jacksonville, Kenner, Marietta, Sandy Springs, Charlotte, Raleigh, and Virginia Beach all exhibit the above mentioned patterns. However, Atlanta, Nashville, and Roswell do not. For example, most of Jacksonville s Latino businesses are not located in block groups where Latinos are concentrated. They are far enough away to produce a separation between residential and commercial districts. In contrast, Nashville and Roswell demonstrate a greater degree of overlap between Latino business and residential areas, while Atlanta shows a tendency towards concentration, but not overlap, of Latino businesses and residences in one part of the city. 40

52 Figure 3-2: Latino population location quotients and business locations by block group to identify clustering. Source: 2000 U.S. Census. 41

53 Figure 3-2: Continued. Source: 2000 U.S. Census. 42

54 Figure 3-2: Continued. Source: 2000 U.S. Census. 43

55 Figure 3-2: Continued. Source: 2000 U.S. Census. 44

56 Charlotte, Raleigh, and Virginia Beach s Latino residences and businesses are dispersed. In Charlotte, once again, there are three distinct areas of city where Latino residences and businesses are concentrated. These areas are in the northern, eastern, and southern parts of the city, located in suburban areas surrounding the urban core. The east sector of Charlotte has the highest concentrations of Latino residences and businesses. However, my ground-truthing of the Central Avenue corridor suggests it is not an enclave. Central Avenue begins near uptown Charlotte and continues eastward through aging suburban neighborhoods with many small strip shopping centers dispersed along the way. For nearly six miles, the corridor is a mix of single family residences, apartments, freestanding commercial buildings, small strip centers, and an aging, enclosed shopping mall near the periphery. Latino businesses along Central Avenue are scattered throughout this area. Similar to Charlotte, Virginia Beach s Latino businesses are located in several areas of city. Most of the Latino businesses are situated along a major east-west commercial corridor, Virginia Beach Boulevard, and along two north-south suburban thoroughfares, Independence Boulevard and Lynnhaven Parkway. Again, the high Latino location quotient block groups are scattered throughout the city. Raleigh s Latino residences and businesses are the most dispersed of any city. Latino block groups and businesses are located in every area of the city except the extreme north and west. Most of the Latino businesses are located on or near major roads in suburban areas. In sum, it would be difficult to argue an enclave is forming here. 45

57 Kenner, in contrast, has two neighborhoods in the north part of the city where all the high Latino location quotient block groups and most Latino businesses are located. Although Kenner has the residential and commercial density of an urban area, it is actually a suburb of New Orleans. Kenner is made up mostly of single-family homes and strip commercial shopping centers, and is the location of the New Orleans airport, a regional hospital, a regional shopping mall, and a golf course. The city is approximately three miles wide by five miles long, or 15 square miles in area. Given these dimensions and concentrated land-use patterns, one might expect Latinos to be more concentrated in Kenner than they in fact are. The city of Marietta is also similar to the other cities as the Latino businesses and the high Latino location quotient block groups are generally dispersed and concentrated in the suburbs. However, Marietta does have some Latino residences and businesses located in the central part of the city. Most Latino businesses in Marietta are located within, or in close proximity to, high location quotient block groups. In Sandy Springs, most of the high Latino location quotient block groups and business locations sprawl along the major commercial corridor. There are two residential clusters of Latinos in Sandy Springs: one in the south and the other in the north part of the city. The Latino businesses are scattered along the major thoroughfare, Roswell Road, and are not necessarily located in the high location quotient block groups. Again, Latino residences and businesses are separated. Although Atlanta s Latino residences and businesses are dispersed throughout the city, the most concentrated area of Latino settlement is in the northeastern part of the city. Arguably an enclave is developing there as this area contains most of the high Latino 46

58 location quotient block groups in the city. Latino businesses are also well represented in this area adjacent to what is arguably an ethnic enclave located northeast of Atlanta s city limits in North Atlanta CDP. There are also many scattered high location quotient block groups in the east-central to southeast part of city, as well as the northwest and west sectors. Generally, the Latino population and businesses are dispersed in all northern and all eastern urban to suburban areas of the city. Also, if Atlanta, Marietta, Roswell, and Sandy Springs are examined as a single area, the results are overwhelmingly congruent: Latinos are dispersed in suburban neighborhoods. Similar to Atlanta, Nashville shows some concentration of Latino residences and businesses in one part of the city. Nashville has a very distinct concentration of high location quotient block groups in the east and southeast parts of the city, with most Latino residences and businesses located in the latter area. An argument could be made that the southeast sector of Nashville is an ethnic enclave. However, given the area and distances over which Latino residences and businesses are spread, it seems a stretch to regard this as a contiguous, cohesive community. It is however possible that several smaller enclaves are forming in close proximity to each other. Furthermore, the north, northeast, and west parts of the city have smaller concentrations that could be developing into ethnic neighborhoods as well. Like in other cities, the Latino businesses locations are spread out along major thoroughfares throughout city, most notably along Nolensville Pike and Murfreesborough Pike. In particular, the northern section of Nolensville Pike contains many Latino businesses (See Figure 3-3) that Mike Davis affectionately refers to as one of the newly developing little Mexicos (Davis 2001: 4-5). Lastly, the most visible concentration of high Latino location quotient block 47

59 groups and businesses is found in Roswell, Georgia. The high location quotient block groups and businesses are somewhat centrally situated. The block groups are located along or intersected by the two major commercial corridors (Roswell/Atlanta/Alpharetta Road and Holcombe Bridge) that form a crossroads near the core of the city. Likewise, most of the Latino businesses are located along these corridors. However, to state that Roswell is developing or has a Latino ethnic enclave may be misleading since it also is an aging suburban city and all development has followed a sprawling linear pattern, not just Figure 3-3: Latino businesses in redeveloped strip center on Nolensville Pike in Nashville, Tennessee. Source: Kristian Dennis. 48

60 that of Latino residences and businesses. Due to Roswell s suburban situation, low Latino block group location quotients, and proximity to Atlanta, Marietta, and Sandy Springs, I contend that Roswell is not a developing Latino enclave. In addition to the previous maps, I created a table that quantifies Latino residence and business clustering (See Table 3-1). The purpose of this table is to show for each city the percentage of Latino businesses located within the boundaries of block groups with high Latino residential location quotients. These percentages definitely reveal that most Latino businesses locate near the Latino population. Table 3-1: Latino businesses located in block groups with a Latino residential location quotient greater than 1.0. Source: 2000 U.S. Census and Kristian Dennis. Businesses in Percent in City Block Groups Total Businesses Block Group Nashville % Roswell % Charlotte % Virginia Beach % Marietta % Raleigh % Kenner % Atlanta % Sandy Springs % Jacksonville % Totals / Average % % 49

61 In 8 out of 10 cities, a majority of Latino businesses are located in high location quotient block groups. Nashville and Raleigh demonstrate especially high concentrations, with 85 and 84 percent of businesses respectively, located in these block groups. In the two cities with lower percentages, the remaining businesses are still within a short distance of those block groups. In all cities, an average of 71 percent of Latino businesses are located in high location quotient block groups. Again, this reveals Latino businesses do tend to cluster in or near areas with significant Latino populations, i.e. high location quotient block groups, regardless of whether the total Latino population is concentrated or dispersed. In sum, this analysis to some degrees confirms heterolocalism s second hypothesis that there is also a lack of spatial overlap between residence on one hand and shopping districts and sites of social activity on the other. Latino residential and business locations do overlap, but they overlap throughout the suburban landscape in what is an arguably heterolocal pattern. 50

62 Chapter 4 Spatial Analysis of Residences and Business Analysis of the Latino business locations and block groups with above average Latino population further validates that Latino businesses do tend to cluster near the Latino population though this population is dispersed. GIS analysis can be used to determine the degree of overlap between the Latino business locations and population. The following maps were created to show the relationship between the block groups with higher concentrations of Latinos and the proximity of Latino businesses to residences (See Figure 4-1). Utilizing GIS, the business locations were geocoded by address to pinpoint their location. Next, I calculated the average Latino block group population for each city and selected the block groups with above average Latino population. The above average block groups were then categorized, using Jenks natural breaks, into 3 categories based on the raw Latino population numbers. The three categories represent average to slightly above average, moderately above average, and highest above average Latino population. Next, I buffered the selected block groups by 1, 2, and 5 miles to identify the Latino businesses within each distance. The businesses are represented on the maps by color-coded symbols (dots) based on the distance each business is located away from a high Latino concentration (above average) block group. The symbols are categorized based on the hot to cold spot (red to light blue) color sequence where businesses located inside the block groups are hot red dots, while businesses located the furthest away from residences are cold light blue 51

63 Figure 4-1: Medium and Large city maps showing high Latino population by block group and Latino businesses by distance from those block groups. Source: 2000 U.S. Census. 52

64 Figure 4-1: Continued. Source: 2000 U.S. Census. 53

65 Figure 4-1: Continued. Source: 2000 U.S. Census. 54

66 Figure 4-1: Continued. Source: 2000 U.S. Census. 55

67 dots. This methodology provides a striking visualization of the relationship between Latino residences and businesses. The maps reveal that Latinos are living in various areas of these cities, even in Roswell, Nashville, Sandy Springs, and Atlanta, because there are multiple block groups with above average Latino population in different parts of the cities. Likewise, Latino businesses are dispersed since many Latino businesses are located inside or near these dispersed block groups. Furthermore, calculating the percent of businesses within the specified distances reveals the extent to which there overlap between residence, shopping, and social sites (See Table 4-1). In this analysis, only businesses located within each city s boundaries were included. For Atlanta, Marietta, Roswell, and Sandy Springs, there were numerous businesses located within the buffer distances but in another city s limits or in unincorporated areas within the county. On average, 60 percent of Latino businesses are located inside above average Latino block groups, while 94 percent of businesses are located within one mile. Lastly, nearly all (99%) businesses are within two miles and all businesses are within five miles of an above average Latino block group. Due to the high percentages of Latino businesses located inside or within one mile of these block groups, it is very difficult to argue there is much separation of Latino residences, businesses, shopping, and social sites. Though I do not analyze employment locations outside of Latino businesses, Ellis et al s (2004) work is illustrative of the national trend: in Los Angeles, the location of Mexican immigrants is dramatically transformed during the workday as they labor in virtually all parts of the metropolitan area. Heavier work tract concentrations do appear in 56

68 Table 4-1: Latino businesses by distance (A) inside, (B) 1 mile, (C) 2 miles, and (D) 5 miles from block groups with above average Latino population. Source: 2000 U.S. Census and Kristian Dennis. (A) Mean Latino Block Group Population Total Businesses In City Limits Businesses Inside Block Group Percent Inside Block Group Medium City Kenner % Roswell % Marietta % Sandy Springs % Jacksonville % Large City Nashville % Charlotte % Raleigh % Atlanta % Virginia Beach % (B) Mean Latino Block Group Population Total Businesses In City Limits Businesses Within 1 Mile Percent Within 1 Mile Medium City Sandy Springs % Marietta % Kenner % Roswell % Jacksonville % Large City Nashville % Atlanta % Virginia Beach % Raleigh % Charlotte % 57

69 Table 4-1: Continued. Source: 2000 U.S. Census and Kristian Dennis. (C) Mean Latino Block Group Population Total Businesses In City Limits Businesses Within 2 Miles Percent Within 2 Miles Medium City Jacksonville % Kenner % Sandy Springs % Marietta % Roswell % Large City Charlotte % Nashville % Raleigh % Virginia Beach % Atlanta % (D) Mean Latino Block Group Population Total Businesses In City Limits Businesses Within 5 Miles Percent Within 5 Miles Medium City Jacksonville % Kenner % Marietta % Roswell % Sandy Springs % Large City Atlanta % Charlotte % Nashville % Raleigh % Virginia Beach % 58

70 East Los Angeles, Santa Ana, the San Fernando Valley, and Ventura County, but the overall impression is of dispersion. No doubt this pattern derives from Mexican immigrant employment in a range of service-oriented jobs that are spread throughout the region (Ellis et al 2004:626). Further, many Latinos work in construction, agriculture, and manufacturing (Kochlar et al 2005:23-24) and these jobs tend to be geographically dispersed inside and outside of cities, I argue this supports the second hypothesis regarding separation of residence and workplace. However, there is overlap between residences and shopping or social sites because both are spread across suburban landscapes. A given Latino block group is likely to have a Latino business or two, but not to have the majority of a city s businesses in the vicinity. The city with the highest percentage of Latino business to residential area overlap is Nashville, where 87 percent of Latino businesses are located within block groups with above average Latino populations. Nashville has the highest concentration of population and businesses in one part of town, even though the southeast area spreads out over 10 miles, across both inner-city urban and suburban landscapes. My analysis reveals several clusters of block groups with above average Latino population in this part of the city. Most Latino businesses in Nashville are located along the three commercial corridors that intersect these areas. However, there are small clusters of Latino residences and businesses developing in most other parts of Nashville as well. Therefore, the structure of Nashville s Latino population and businesses is predominantly a dispersed, suburban pattern with potentially some ethnic enclave formation. Similarly, Atlanta, Sandy Springs, and Roswell show some concentrations of 59

71 Latino population in various parts of the city. However, the degree of overlap by Latino businesses is much closer to the 60 percent average overlap of Latino businesses for all cities in this analysis. This analysis indicates that many Latino businesses are located on the periphery of block groups with above average Latino populations but that these block groups are themselves scattered. In most cities, Latino residences and businesses are generally dispersed and do not reach high levels of concentration. Maps identifying block groups where the Latino population is overrepresented reveal only limited overlap between block groups with the highest Latino population and business locations. The block groups with the highest Latino population contain minimal percentages of businesses and average 6.6 percent of Latino businesses overall (See Table 4-2). This lack of overlap may be due to the dispersed, suburban settlement patterns where most residential and commercial areas are separated OR attributed to how the block group boundaries were drawn. Consequently, I contend based on these analyses that Latino residences, shopping, and social sites are not generally separated and there is only a lack of significant overlap in the cities where Latinos are most concentrated. 60

72 Table 4-2: Percentage of businesses located in block groups with the highest Latino populations. Source: 2000 U.S. Census Bureau and Kristian Dennis. City Number of Highest Population Block Groups Number of Latino Businesses in Block Group Percent of Total Latino Businesses Jacksonville % Kenner % Marietta % Roswell % Sandy Springs % Atlanta % Charlotte % Nashville % Raleigh % Virginia Beach % Average % 61

73 Chapter 5 Assessment of Latino Community Ties Survey Methodology The last major component of my analysis involves testing Zelinsky s third hypothesis of heterolocalism, which states despite the absence of spatial propinquity, strong ethnic community ties are maintained via telecommunications, visits, and other methods at the metropolitan, regional, national, and even international scale (Zelinsky 2001: 133). In order to test this hypothesis, I surveyed customers and business owners at a random sample of local tiendas (grocery/food and retail stores) and restaurants in the cities where Latinos are most concentrated to determine whether or not Latino community ties are strong and if they are, how those community ties are maintained. I was interested in looking at the variety of ways Latinos stay in touch with their culture and community and determining whether it is via telecommunications, personal visits, community events, etc. Though person-to-person contact remains critical in ethnic community formation, I explore the variety of ways in which telecommunications and media technology augments these ties. The survey questions I developed measure community ties by assessing social, religious, media, recreational, residential, and business activities (See Table 5-1 ). After developing the survey questions, I had to determine where to conduct the surveys. First, I combined my Latino business tables compiled in the GIS analysis for all 15 cities to create a complete list of businesses. Then, I used Research Randomizer s 62

74 Table 5-1: Survey results of how Latino community ties are maintained. Source: Kristian Dennis, October Survey Question How often do you socialize with other Latinos? Do you attend religious services? Do you read Latino newspapers Do you listen to Latino radio? Response Responses (out of 41) Percent of Total Several times a week 11 27% Once a week 12 29% 2-3 times/month 6 14% Once per month 4 10% Less than once per month 8 20% Yes 21 51% No 20 49% Yes 30 73% No 11 27% Yes 29 71% No 12 29% Are you a member of recreational or Yes 5 12% soccer club? No 36 88% Are there other ways you maintain contact with the Latino community? Why do you live where you live? Yes: Work 12 29% Other 3 7% Reunion-family or social 2 5% School 1 2% Nightclub/billiards 1 2% Telephone 1 2% No: 21 51% Latino community 4 10% Small town 5 12% Religious work 1 2% Good for kids 1 2% Other Family Nearby 4 10% Nice city 2 5% Quiet 5 12% Work 5 12% N/A 14 34% 63

75 Table 5-1: Continued. Source: Kristian Dennis, October Survey Question Why did you locate your business where you did? Response Responses (out of 41) Percent of Total Near other or larger stores 3 7% Near where Latinos live 7 17% In city 2 5% Family 2 5% N/A 27 66% 64

76 random number generator to select 45 out of the 818 Latino business locations to target for field research. While conducting the field research, if one of the selected locations was closed or there were no survey participants, I selected another location nearby. In all, I was able to conduct 41 complete Latino community surveys. Survey Discussion First, the frequency and amount of social activity that surveyed Latinos participate is high. In fact, 56 percent of Latinos surveyed socialize once per week or more, while 80 percent of Latinos surveyed socialize at least once per month or more. Secondly, Latino newspaper readership is strong with 73 percent of Latinos reading them regularly. This readership indicates Latinos are keeping up with Latino community events and news. Similarly, radio listenership is strong with 71 percent of Latinos regularly listening to Latino radio, which also indicates Latinos are keeping up with music, local news and events. Furthermore, religious participation is solid with slightly over half of Latinos surveyed attending services. Traditionally, attending religious services has been a way for people to build, maintain, and strengthen community ties (Zelinsky 2001:7). Perhaps the best indicator of how strong community ties are, and how they are maintained, is the diverse responses given for other ways Latinos maintain interpersonal contact and communication with the Latino community. Not surprisingly, work is the primary method Latinos use for contact with other Latinos. However, school, nightclubs and billiards, family and social reunions, telephone, and undisclosed other are also listed as ways Latinos maintain community ties. Based on the number of restaurants, nightclubs, and billiards places in the medium and large cities, these places are likely primary socialization locations. 65

77 Similarly, Latinos responded with equally diverse answers when asked why do you live where you do? The answers to this question further explain and support the locational analyses in the previous sections. Responses by Latinos are centered on various quality of life characteristics that make day-to-day living in a community appealing. In fact, Latinos are living in different places for other reasons than just for work or to be near other Latinos. Only 12% of Latinos cited work and 10% cited being near other Latinos as the primary reason in deciding where to live. Conversely, characteristics such as a city being a small town or being quiet and peaceful or being a nice city are listed by 29% of Latinos surveyed as the primary reason for living in a particular city. Likewise, a city being a good place to raise children, the presence of other family members, and religious work are also factors Latinos considered when deciding where to live. As for business locations, Latino business owners provided some similar answers about why they chose their business location, such as being located in the city or because other family lives there as well. However, the most common responses provided by Latino businesses owners were that they had chosen to locate close to their customers. Only half of all Latino business owners cited being near other Latinos as the primary factor in deciding location, while 50 or 100 years ago this would have been 100 percent. Additionally, several business owners (21%) cited being near other stores of any type or near larger stores, such as medium-sized and big box retailers, as the deciding factor in selecting location. One store owner I interviewed in downtown Siler City, NC actually told me his success story of coming to work in the United States and eventually starting his own business that sells dry goods, men s and women s clothing, western wear, shoes 66

78 and boots, music and musical instruments, and jewelry: I came here in 1995 with my brother to work tobacco and pigs. We were by ourselves, the only two (Latinos) here and couldn t speak English. We worked as much as we could and saved as much money as we could, so we could get our families here. We liked the small town so we stayed. (As) more and more Latinos came to work the farms, I decided to open a store to serve our (growing) community. It s been open for six years. This testimonial, along with the various other survey responses, further proves that Latino community ties are strong, are maintained a variety of ways in the absence of significant spatial clustering, and physical contact and communication is important. Based on these results, I can conclude that Latino community ties are strong but that businesses still cite being located close to their Latino customer base as important. Latino Media Evaluation Additional field research in these cities provides information about the extent and range of Latino media outlets; such as Latino newspapers, radio stations, satellite and local TV stations, and local internet websites. Generally, the growth and progression of Latino media from small to large cities follows a sequence. Typically, satellite television service is available first, followed by Latino newspapers, then radio stations, magazines, websites, and lastly, local Latino television programming or stations (See Table 5-2). However, this progression does not appear to limit the scope of each Latino media. Latino media in new destination cities in the South serve neighborhood, local, regional, national and international markets. For example, northern Sandy Springs and southern Roswell have a developing Brazilian community where a Portuguese newspaper (The World Journal), a Brazil/Atlanta newspaper (Jornal Moderno), and a magazine (Cia' Brasil) are available at local stores. Likewise, the entire Atlanta metro area, including 67

79 Table 5-2: Latino media by city and category. Source: Kristian Dennis, October Bells, TN City Small City Newspapers and Radio / TV Stations Magazines 1520 AM Latinos News- Memphis Website Biscoe, NC various Piedmont Que Pasa-Piedmont Que Pasa-Piedmont quepasamedia.com Collinsville, AL Ft. Payne- Gadsden Latino News Paisano Alabama latino-news.com paisanonewspaper.com Siler City, NC see Raleigh 1530 AM Que Pasa radio Que Pasa-Piedmont quepasamedia.com see Charlotte WXNC 1060 AM La Tremenda Monroe, NC Que Pasa-NC/Charlotte quepasamedia.com El Soplon Deportivo- elsoplondeportivo.com also available in NC, SC and Charlotte Medium City 880 am Jacksonville, NC 910 AM La Invasora 1180 AM Que Pasa-NC quepasamedia.com 830 AM La Fabulosa Radio Tropical KGLA 1540 AM Kenner, LA Sur TV 55 La Prensa Nuevo Orleans Jambalayo Deportivo Vocero voceronews.com 68

80 Table 5-2: Continued. Source: Kristian Dennis, October Newspapers and Medium City Radio / TV Stations Magazines Website Marietta, GA see Atlanta Roswell, GA see Atlanta Sandy Springs, GA see Atlanta Large City RadioMex 610 AM radiomex610atlanta.com 980AM Radio Lider 1030 AM Radio Fortaleza 1040 AM 1080 AM 1100 AM 1190 AM 1340 AM ESPN Deportes 1420 AM 1460 AM1600 AM Atlanta, GA CNN Radio Noticias Planet X 96.5 FM WAZX 1550 AM/ radiolaquebuena.com FM La Raza laraza1023.com Viva FM vivaatlanta.com Azteca TV 4 Telemundo 38 Nuevo Impacto nuevoimpacto.net Vida Latina vidalatina.cc Georgia Latino News galatinonews.com Impacto Latino impactolatino.net Pasos Para Vivir Mejor mundohispanico.com 69

81 Table 5-2: Continued. Source: Kristian Dennis, October Newspapers and Large City Radio / TV Stations Magazines Website Competencia: Revista revistacompetencia.com Latina de Negocios Estadio Deportes estadiosports.com Accion Mundo Hispanico mundohispanico.com (Georgia) Mas masatl.com Atlanta Latino atlantalatino.com Perez Montibelli perezmontalbetti.com Atlanta, GA SDR Inc. sdrnet.com Atlanta La Vision lavisionnewspaper.com Cia Brasil Jornal Moderno jornalmoderno.net Atlanta/Brasil The World Journal twjournal.com El Progreso Hispano- elprogresohispano.com GA, NC, SC, NY, NJ Latino USA latinousainc.com Que Pasa Radio Que Pasa-NC quepasamedia.com WBZK 980 AM La Tremenda: 1060AM /1310 AM/102.3 FM Charlotte, NC WNOW AM 1030 Que Pasa- Charlotte quepasamedia.com Mi Gente- NC, SC, migenteweb.com Charlotte 70

82 Table 5-2: Continued. Source: Kristian Dennis, October Newspapers and Large City Radio / TV Stations Magazines Website Reven- Charlotte and quepasamedia.com metro NC cities. La Noticia El Sol lanoticia.com Charlotte, NC Vida Latina Estadio Lider En vidalatina.cc estadiosports.com Deportes Latino USA latinousainc.com El Progreso Hispano elprogresohispano.com La Sabrosita 810 AM 900 AM 1130 AM Activa1240 AM Activa1240am.com 1300 AM Telefutura TV 42 TBN Enlace 51 Nashville, TN La Campana-Nashville Latino News lacampana.us El Crucero De latino-news.com Tennessee La Noticia La Voz De Nashville hispanicpaper.com Latino USA latinousainc.com hispanicnashville.com 88.1 AM 96.9 FM La Ley Raleigh, NC 540 AM 1000 AM 1390 AM 1530 AM 71

83 Table 5-2: Continued. Source: Kristian Dennis, October Newspapers and Large City Radio / TV Stations Magazines Website 88.1 AM 96.9 FM La Ley 540 AM 1000 AM 1390 AM 1530 AM Que Pasa Radio 1310 quepasamedia.com AM Raleigh, NC WZGS Telemundo- T44 Telefutura 26-(Durham) The Bilingual thebilingualus.com Que Pasa-NC quepasamedia.com La Conexion- Raleigh laconexionusa.com Que Pasa- Triangle La Noticia Weekly quepasamedia.com Latino USA lanoticia.com latinousainc.com Selecta 1050 AM El Eco de Virginia (Norfolk) Tidewater Hispanic Virginia Beach, VA La Voz Hispana de Virginia 72

84 Marietta, Roswell, and Sandy Springs has numerous Latino media available ranging from local to international in scope. The diversity and accessibility of Latino media is also evident in North Carolina, where Biscoe, Siler City, Jacksonville, Charlotte, and Raleigh have local, metropolitan, state, regional, domestic, and international media available. Similarly, Nashville s Latino media presence is strong and diverse, while Kenner and Virginia Beach have solid Latino media markets with local and regional media available as well. Latino Business Climate I combined the Latino business GIS table and the Latino media table to provide a quantitative summary of types of businesses in each city (See Table 5-3). The smaller cites mainly have Latino restaurants and grocery stores. As a city and the Latino population increase in size, so does the availability of professional businesses, services, and media. However, the Latino economy in several small and medium cities is diversifying as well. A great example of this diversification can be found in Collinsville, Alabama. Collinsville is the smallest city in this study, yet contains 13 Latino businesses distributed among the grocery, miscellaneous, professional, restaurant, and retail categories. Equally impressive is that the Collinsville city government s website has many of the city s Latino businesses listed in the business directory. This fact demonstrates the city of Collinsville is proactive in promoting its economy, while at the same time rejuvenating its central business district (CBD). In fact, most cities in this study, especially the smaller ones, have benefited from Latino revitalization of downtown areas (See Figure 5-1). 73

85 Table 5-3: Latino businesses and services within city limits by category with population and dissimilarity indices (DI). Source: 2000 U.S. Census and Kristian Dennis, October Small Cities Bells Biscoe Collinsville Monroe Siler City Population 2,171 / 1,700 / 1,644 / 26,228 / 6,966 / Total / Latino ,611 2,740 ( % ) ( 22.8) ( 23.2 ) ( 23.5 ) ( 21.4 ) ( 39.4 ) Population DI Grocery Media 1 Misc.* 5 3 Prof.** Religion 2 3 Restaurant Retail Total Medium Cities Jacksonville Kenner Marietta Roswell Sandy Springs Population Total / Latino ( % ) 66,715 / 6,702 ( 10.0 ) 70,517 / 9,602 ( 13.6 ) 58,748 / 9,947 ( 16.9 ) 79,334 / 8,421 ( 10.6 ) 85,781 / 8,514 ( 9.9 ) Population DI Grocery Media Misc.* Prof.** Religion Restaurant Retail Total

86 Table 5-3: Continued. Source: 2000 U.S. Census and Kristian Dennis, October Large Cities Atlanta Charlotte Nashville Raleigh Population 416,474 / 540,828 / 545,524 / 276,093 / Total / Latino 18,720 39,800 25,774 19,308 ( % ) ( 4.5 ) ( 7.4 ) ( 4.7 ) ( 7.0 ) Virginia Beach 425,257 / 17,770 ( 4.2 ) Population DI Grocery Media Misc.* Prof.** Religion Restaurant Retail Total * Miscellaneous includes all businesses or services not in other categories. ** Professional services include F.I.R.E., Legal, Medical, etc... 75

87 Latino businesses on Main Street. Figure 5-1: Revitalization of Collinsville, Alabama CBD due to Latino economic growth. Source: Kristian Dennis, October